| With the rapid development of computer science and technology,3D model has been widely used in virtual reality, cultural relic archeology, medical imaging systems, film, television, video games and other areas. Compared with2D images, it is closer to the reality, more consistent with human visual habits and can produce more vivid effects. The amount of3D data grows rapidly along with the development of3D acquirement technique, thus there are demands on computer storage, graphics rendering, processing, network transmission and so on. It should be noted that in some cases we do not require the high-resolution details of the model. Therefore, based on the analysis of3D model acquirement technique, we adopt the one which uses binocular stereo vision system based on Bumblebee2. Moreover, the theory of simplification on3D mesh model is elaborated, as well as the analysis of advantages and disadvantages of the proposed algorithms. Then, we propose a new algorithm to simplify the3D mesh model and achieve a good compromise between massive3D data and computation workload.In this thesis, we first introduce the background, significance of our research and the research progress on acquirement technique in3D model and3D mesh model simplification. The discussion of binocular stereo vision in3D reconstruction and the critical processing steps including camera calibration, image preprocessing, stereo matching and3D reconstruction, etc. are also given. We describe each step of getting and processing3D image data including3D reconstruction forward by using Bumblebee2binocular camera in detail. Furthermore, we analyze shortcomings of the QEM classical algorithm in the process of3D mesh model simplification and present an edge collapse mesh simplification algorithm based on the roughness. By defining roughness based on vector of vertex as weights and embedding quadric error metric in QEM algorithm, the error measure can measure distance deviation and reflect the important geometric details of the model local surface characteristics. After the large-scale simplification, it can still keep important details of model characteristics.Based on the proposed work, we develop a software system to simplify the3D mesh model by using OpenGL. By using of the simplified mesh human face model and rabbit mesh model, we compare the proposed algorithm with QEM algorithm and its improved algorithm. Experimental results verify the efficiency of the proposed algorithm. |
